HEADLESS APP
Knowledge Server for Agents and Apps
Run Lorebo on a server you keep online around the clock. Agents and applications can read and update your knowledge through built-in tools and SDKs, even when your laptop is off.
WHY HEADLESS
Knowledge Shouldn't Depend On One Laptop
Give your agents, applications, and colleagues a shared source beyond your personal computer.
WHEN YOUR LAPTOP GOES OFFLINE
Work Continues. Access Stops.
A base on your laptop is available only while the computer is awake, connected, and running Lorebo. Background work and colleagues lose access when it goes offline.
WHEN TOOLS KEEP SEPARATE COPIES
Separate Copies Drift Apart
An agent updates a decision, but an application still reads an older copy. Separate copies leave people and tools working with different versions of the same knowledge.
ONE BASE ON YOUR INFRASTRUCTURE
Keep the Source Where Work Can Reach It
Run Headless on a server you keep online around the clock, so access to knowledge does not depend on your laptop.
Agents and applications read and update the same hosted base, with the access you grant them.
Connect Desktop for visual reading, editing, and graph inspection.
Video coming soon
We’re preparing a walkthrough for this section.
GET UP AND RUNNING
Deploy. Connect. Extend.
DEPLOY YOUR RUNTIME
Start Your Runtime
Install a native build or run Docker on a supported host.
Keep the service running and its data persistent. Installation Options
CONNECT YOUR TOOLS
Connect Your AI
Follow Get Started to connect your AI client through MCP.
Local MCP works on Free. Use the CLI to administer the service. MCP Setup
EXTEND WITH PRO
Add Pro Features
Choose Pro through your account when you need paid capabilities.
Activate the runtime with lorebo-headless access activate. Compare Plans
QUICK START
Install and Run Headless Your Way
Choose an installer or container, then operate your service from the command line.
Install the native runtime on your computer or server, initialise a profile, and start the service from your terminal.
curl -fsSL https://downloads.lorebo.com/headless/install.py | python3 -
~/.local/bin/lorebo-headless init
~/.local/bin/lorebo-headless serve-httpInvoke-RestMethod https://downloads.lorebo.com/headless/install.ps1 | Invoke-Expression
& "$HOME\.lorebo\bin\lorebo-headless.cmd" init
& "$HOME\.lorebo\bin\lorebo-headless.cmd" serve-httpYour local service endpoint
HTTP and MCP listen on 127.0.0.1:8758 by default. The platform downloads below provide a native build as an alternative to the installer.
Run the official image with persistent storage and a local port.
docker run --detach --name lorebo \
--publish 127.0.0.1:8758:8758 \
--volume lorebo-data:/data \
ghcr.io/lexrag/lorebo-headless:stableKeep data across container updates
The lorebo-data volume stores your profile at /data. Back it up before upgrading; pin an image version or digest for repeatable deployments.
Check service health and manage pages from your shell.
lorebo-headless status
lorebo-headless doctor --network
lorebo-headless page list
lorebo-headless --helpKnow which checks need the service
status works without a running server. doctor --network and page list need the HTTP service. Use MCP or HTTP for broader graph operations.
Download
macOS
Command-line app for Apple silicon Macs and local automation
Checking download availability…
Download
Windows
Command-line app for Windows services and agent hosts
Checking download availability…
Download
Linux
Command-line app for Linux servers, containers, and agent hosts
Checking download availability…
PYTHON OR TYPESCRIPT SDK
Give Your Applications Access to Your Knowledge
Save knowledge and get answers from your Lorebo base directly in your application. Use Python or TypeScript with the model of your choice.
CAPTURE
Capture Knowledge
Send useful text to capture. The SDK searches existing knowledge first.
It updates pages that cover the topic and creates pages for new knowledge.
EXPLORE
Explore Knowledge
Send a question to explore. Receive an answer with source references.
Your application can inspect the referenced pages when it needs the full evidence.
INTEGRATE
Use Your Stack
Connect your Python or TypeScript application to a running Lorebo service.
Your app controls the workflow. Your model receives the retrieved knowledge.
EASY INTEGRATION
A Knowledge Service That Fits Your Applications
Run Lorebo on your own server and connect it to the agents and applications you build.
FLEXIBLE INSTALLATION
Run It Where You Need It
Install Lorebo with a native build or the official Docker image on a server you control and maintain.
Keep knowledge available to your agents and applications even when your personal laptop is off.
BUILT-IN CONNECTIONS
Connect the Way You Build
Connect agents through built-in MCP tools and applications through the Python or TypeScript SDK.
Capture knowledge and get answers from the hosted base using the model and workflow you choose.
KNOWLEDGE FOR YOUR APPLICATIONS
You Build the App. Lorebo Handles the Knowledge.
Lorebo brings the base, knowledge tools, and application access together in a service you run on your own server. Your agents and apps can read and update the same hosted knowledge.
Your application keeps control of its workflow and model. Lorebo supplies the knowledge service, so you can focus on what your application does with that knowledge.
Video coming soon
We’re preparing a walkthrough for this section.
Frequently Asked Questions
Good to Know
Do I need a Lorebo account?
No. Free Headless starts without account activation. Activate the runtime when you need Pro capabilities, including base sharing and working across several bases.
How do I keep it running?
Use your operating system’s service manager or a container restart policy to start Headless after a reboot. Confirm the service is healthy before reconnecting clients.
What should I back up?
Copy the complete profile and every hosted base, including attachments, with Headless stopped. For Docker, preserve the data volume and record the image version for recovery.
How do I connect remotely?
Use your own protected endpoint for applications. Compatible AI clients can also use Lorebo-managed remote access with Pro. Keep Headless running on the server.
Can several clients connect?
Yes. Agents and apps can use one running service with the base access they need. Keep one runtime in charge of the profile; do not start two writers on the same files.
Where does the AI model run?
Headless serves the knowledge base. Your agent or app chooses a local or external model. External AI services receive the knowledge included in the requests you send to them.
Explore More

Copy the setup commands for your target system to install, initialise, and start Lorebo on your own machine or cloud server.